AMD: Radeon VII Has Excellent Results with DirectML; We Could Try a GPGPU Approach for Something NVIDIA DLSS-like
One of the jabs directed by NVIDIA CEO Jensen Huang to the AMD Radeon VII reveal at CES 2019 is that the upcoming GPU doesn’t have any hardware support for raytracing or AI-based super-sampling like DLSS, unlike the NVIDIA RTX line of graphics cards.
In an interview with Japanese website 4Gamer.net, Adam Kozak (Senior Manager of GPU Product Marketing at AMD) addressed both topics. The most interesting part is that Radeon VII is already performing well through Microsoft’s DirectML API, according to Kozak, and he also suggested AMD could even try to develop something similar to NVIDIA’s Deep Learning Super-Sampling with a GPGPU approach instead.
At last year’s Game Developers Conference 2018, Microsoft announced a framework “Windows ML” for developing machine learning based applications on the Windows 10 platform, and “DirectML” that makes it available from DirectX12.
We are currently experimenting with the preview version SDK of DirectML, but Radeon VII shows excellent results so far.
By the way, Radeon VII scored about 1.62 times the GeForce RTX 2080 in “Luxmark” which utilizes an OpenCL-based GPGPU-like ray tracing renderer.
Based on these facts, I think NVIDIA’s DLSS-like thing can be done with a GPGPU-like approach for our GPU.
DirectML is currently due to be available in Spring 2019. We actually reached out to Microsoft a while ago and received the following statement regarding its extensive capabilities:
DirectML provides a DirectX 12-style API that was designed to integrate well into rendering engines. By providing both performance and control, DirectML will enable real-time inferencing for game studios that want to implement machine learning techniques and integrate them into their games. These scenarios can include anything from graphics related scenarios, like super-resolution, style-transfer, and denoising, to real-time decision making, leading to smarter NPCs and better animation. Game studios may also use this for internal tooling to help with things like content and art generation. Ultimately, we want to put the power into creators’ hands to deliver the cutting edge experiences gamers want across all of the hardware that gamers have.
AMD’s Adam Kozak also touched on raytracing in the aforementioned interview from CES 2019, clearly stating that there isn’t much point in supporting the technology for the time being due to its software availability, limited to just Battlefield V and the Port Royal benchmark.
I understand the importance of DXR, but I think it is too early to use it in games. Ray tracing is useful for professional video production sites, but there is Radeon Pro Render for such applications.
The only compatible game is Battlefield V, in addition to the Port Royal benchmark. It will be irrelevant to 99% of gamers. It is obvious that there is little benefit for gamers so far.
However, there is no doubt that we will support it. I cannot say anything more at this time.
That’s essentially what AMD CEO Lisa Su said a few days ago with a little more nuance and diplomacy. It may be a while before we learn any concrete details about AMD’s plans for raytracing or AI-based deep learning techniques like NVIDIA’s DLSS, but customers can take solace in knowing that both are being worked on.